Using Agricultural Waste for Remediation of Chemical Contaminants in Water
Part of the Young Naturalist Awards Curriculum Collection.
Introduction
“Water is essential for life. It is a prerequisite for human health and well-being” (“Water for Life”). Yet 11% of the world’s population does not have access to safe drinking water (“Millennium Development Goal”). The Safe Drinking Water Act regulates standards, but it only applies to public water-treatment plants and not to private wells (“Safe Drinking Water Act”). In the United States, more than 15 million households get their drinking water from private wells that are not treated or even monitored for safety (“Private Ground Water Wells”). In my home state of Arizona, 250,000 people fall into this category (Artiola et al., 19). The situation is much worse in developing countries, where access to safe drinking water is very limited. Learning about these issues at a Girl Scout Jamboree made me think how often people take water resources for granted. It inspired me to research chemical contaminants to promote safe drinking water and protect the environment in general.
Many things can contaminate water sources, but this project focused on herbicides, fertilizers, and fungicides, which enter surface water from urban and agricultural runoff and percolate into groundwater from farm irrigation and urban lawn watering (“Managing Stormwater Runoff”). Contaminants are documented at levels above accepted guidelines for safe drinking water (“Agricultural Chemicals”; “Understudied Fungicides”; “Study Confirms Glyphosate”; Biello), and may be carcinogenic or cause other problems for body systems (“Drinking Water Contaminants”).
Public water treatment disinfects and removes sediments and harmful chemicals. Filtration, one step in water treatment, uses activated carbon to remove chemicals from water by binding and removing them, a process called adsorption. Although effective, disadvantages include expensive equipment and frequent replacement of materials (“Water Treatment”; “About Activated Carbon Filtration”). Locally available, low-cost materials as alternatives to activated carbon would be beneficial. Researchers in Taiwan successfully made an activated carbon-like product from rice straw waste and used it as an adsorbent to remove pesticides from water (Chang et al. 348). In Spain, researchers adsorbed pesticides from water using organic wastes from olive, date, and avocado stones, bamboo canes, peanut shells, and sawdust (Bakouri et al. 336). In Wisconsin, pine bark was effectively used in adsorbent columns to remove pesticides (Tshabalala 1). Adsorption can take place at any solid-liquid interface. The solute is the substance being removed, and the adsorbent is the solid material used for the process (Armenante). There are two methods: filtering water through adsorption columns, or batch adsorption, which is a mixing process (“Adsorption/Active Carbon”; Bungay).
The Index of Refraction (IR) is unique and specific for different solutions because “light travels at different speeds in different media [and] …when light passes from one medium to another at any angle other than 90˚, it not only changes speed, it also changes direction at the boundary between the two media.” (“Refractive Index Principle,” 2012) The IR is affected by the concentration of the solution being measured, with IR increasing as the concentration increases. Because temperature affects IR readings, the temperature at the surface of the prism should be kept constant. (K-Patents, 2011) IR was used in this project to measure the concentrations of chemical contaminants in water.
The solution’s concentration can be measured using an Abbe refractometer. The index of refraction (IR) of a liquid is found by passing light through the sample and measuring the angle made when it hits the surface (Figure 1). The IR is specific for a solution, and increases as the concentration increases (the refractive index). By measuring IR for a series of prepared dilutions, a calibration curve can be created for a particular chemical, and the line’s equation calculated. The unknown concentration of other dilutions of that same chemical can be determined by measuring the sample’s IR and solving for concentration in that line’s equation: y=mx+b, where y is IR and x is concentration (slope m and y-intercept b are constants for that line).
This project studied batch adsorption using agricultural wastes found locally in Arizona to evaluate their performance as adsorbents to remove chemical contaminants from water.
Research Plan
In the early stage of my research, I collected water from irrigation ditches alongside farms near my home. I tested IR and conductivity for these agricultural runoff samples before and after batch adsorption, varying the solid-to-liquid ratio, the adsorbent particle size, and the contact time. Results provided trends to understand how such factors affected the adsorption processes, but could not describe what was in the water. This work was the baseline from which I planned research studying how contaminant concentration is affected by batch adsorption.
The research question is: Can agricultural waste be used to remove contaminants from water? The hypotheses state: 1) if water contaminated with a chemical is mixed with agricultural waste in a batch adsorption process, then the concentration of contaminant in the water will be decreased because the waste material is able to bind with and remove the contaminant from the solution, and 2) there is an optimal length of contact time for mixing in batch adsorption that causes the greatest reduction in contaminant concentration.
The dependent variable measured was contaminant concentration in water samples. The independent variables tested were: 1) agricultural waste used as adsorbents (pine bark, palm stem, corncob); 2) contaminant (herbicide, fertilizer, fungicide); and 3) contact time (one, two, three, four hours of mixing on the ball mill roller). A control for each contaminant had no adsorbent added. Based on a literature review, the constant conditions were an approximate 2.5% contaminant concentration, a solid-to-liquid ratio of 20 g adsorbent to 1 L contaminant solution, and adsorbent particle sizes of 0.76 mm to 1.52 mm (Bungay; Armenante). All contaminant samples used a 20 mL starting volume.
Three adsorbents were tested for each of three contaminants, making a total of nine testing conditions. Four mixing (contact) times were evaluated. Three trials were conducted; IR was measured twice for each experimental and control sample. This plan represents 144 samples, with 288 measurements included in data analyses.
Materials and Methods
The Vee-Gee C-10 Abbe refractometer was calibrated (per the instruction manual), using distilled water as the blank. A VWR Model 1125 water circulator with a temperature controller was attached to the refractometer to maintain a constant temperature (27˚C) at the prism surface, which is necessary for IR accuracy (Figures 2A & 2B).
The adsorbent materials were gathered: Pine bark was removed from firewood; palm stems were collected from a Mediterranean fan palm in my backyard and the thorns removed; corn was boiled, the kernels removed, and the cobs dried. Materials were cut into approximately half-inch pieces, spread on sheets in a single layer, and baked in the oven at 225˚F to remove excess moisture (Figures 3A - 3D). After cooling, the materials were ground using a commercial-grade conical burr grinder. Ground materials were sieved to select particle size, first using a Geotech 30 OPN sieve to remove the small particles, then sieving the remaining materials (60 OPN) to discard the larger particles (Figure 4A). The particles left (0.76 mm to 1.52 mm) were used for adsorption tests (Figures 4B & 4C).
For each contaminant, four dilutions were prepared by weight (7.5%, 5%, 2.5%, 1%) using a digital scale. Solutions (10 mL) of Roundup Weed and Grass Killer Herbicide (glyphosate isopropylamine salt 50.2%), Schultz Liquid Plant Fertilizer (10% nitrogen-15% phosphate-10% potash 35%), and Hi-Yield Consan Fungicide (n-alkyl dimethyl benzyl ammonium chlorides 20%) were prepared using the equation X=CA/(D+A), where C is contaminant concentrate, A is contaminant amount, and D is the amount of distilled water. IR readings were taken for each contaminant, and Excel was used to plot IR versus concentration to create a tool to calculate the unknown concentration of the experimental samples.
Stock contaminants were prepared as 2.5% solutions by volume using the equation above, measured with a graduated cylinder, and stored in gallon jugs. Thirty glass bottles (30 mL) with screw caps were filled, following the research plan (Figure 5) and using a digital scale to measure 20 g of contaminant (Figures 6A & 6B) and 0.4 g of adsorbent into each sample. Bottles were loaded into a paint can, and Styrofoam peanuts were used to prevent breaking. The can was sealed and placed on the ball mill roller (Figures 7A & 7B).
The can was opened after one hour, the controls were removed, and a sample was taken from each using a pipette, then returned to the can. Bottles for the one-hour tests were removed, and the can was repacked and returned to the ball mill roller to continue mixing. The one-hour samples were filtered to remove solid adsorbents, and two IR measurements were taken for each sample (Figures 8A & 8B). These steps were repeated with the two-, three-, and four-hour samples.
Results
Adsorbent materials were heat-treated before experimentation to remove excess moisture (Figure 9) so the materials were not holding water in the pores on their surfaces to which contaminants could bind, and also to prevent water from being released into the samples during mixing, which could affect their concentration.
The calibration curve tool (Figure 10) for each contaminant was generated using Excel by plotting IR versus concentration, and also the line equations y=mx+b. For each contaminant, the y-intercept b is close to distilled water (IR=1.3330), showing that the data is accurate, because as it crosses the y-axis, the contaminant concentration would equal zero, as for pure distilled water. The correlation coefficient (R2) confirms the data is reliable because the values are very close to 1.0. Data points were close to the equation line, with no outliers. Results support that the IR measurements were done accurately, and the equation was reliable to use as a tool to calculate concentration.
The starting concentrations of stock contaminant solutions were measured by IR and calculated using the respective calibration curve equation. Starting concentrations were 2.87% herbicide, 2.39% fertilizer, and 2.55% fungicide; all were close to the 2.5% target. These variations were considered when calculating the change in concentration; the project’s data was not affected because each contaminant solution was treated independently, based on its actual starting concentration, and was not assumed to be 2.5%.
The experiments were conducted in my garage, and the ambient temperature was relatively constant during batch adsorption. IR data representing all trials was averaged for each contact time, and graphs were made for pine bark (Figure 11), palm stem and corncob (Figure 12). Results for all testing conditions show that adsorption of contaminants took place using agricultural waste as media, documented by a change in IR readings reflecting reductions in concentration. The control for each contaminant showed no significant change in concentration. For the experimental samples, IR decreased with time to a low point at two hours of mixing, after which it began to rise again. By solving for in the equation for each calibration line, the concentrations of the respective contaminant samples were calculated and graphed to allow comparison of each contaminant/adsorbent combination (Figure 13).
Using the contaminant solution’s concentration at starting time and after batch adsorption, the percentage change in concentration was calculated using Excel [(C2–C1)/C1] x 100; where C1 is thestarting concentration and C2 is theconcentration after mixing. Figure 14 compares testing conditions displaying the percentage reduction in concentration at two hours, which was the contact time achieving the lowest contaminant concentration and therefore the most adsorption. Results show that the combination of herbicide with pine bark had the greatest change in concentration (15% reduction). Overall, pine bark was the most effective adsorbent, showing the highest reductions in concentration across the three contaminants; corncob was least effective. The fungicide was the most easily adsorbed contaminant, showing the highest reductions in concentration across the three adsorbent materials. Fertilizer was the least adsorbed.
Discussion
“Adsorption is a physical separation process in which the adsorbed material [contaminant] is not chemically altered” (Armenante). IR is a reliable testing method because the value is specific to the contaminant’s chemistry, which remains unchanged during the batch adsorption process. This project’s research question was answered positively, and the hypotheses were supported by the data, using reliable experimental methods.
The binding of molecules occurring at the solid-liquid interface in adsorption is different from absorption, which is when molecules get taken into a liquid and the binding forces are strong. In adsorption, the binding of solute molecules to the adsorbent’s surface is usually weak and apt to separate, and a surface that is already full of solute molecules is unable to allow additional binding (“Adsorption/Active Carbon”; Bungay). I believe this is what happened when the contact time was increased beyond two hours—contaminant molecules became unbound from the adsorbent’s surface and returned to the solution, increasing its concentration, which demonstrates that there is an optimal length of mixing time at which the greatest adsorption is achieved.
Differences in an adsorbent’s surface structure, and the size and shape of its surface pores, may have affected how well the solute molecules could bind, or how strong the bonds were. Pine bark was the most effective adsorbent and was the most fibrous, dense material tested in this project. Corncob, which was the least effective adsorbent, was the softest and most porous. The surface area of the adsorbent material also affects how much adsorption can take place. Future work could test the effects of varying the particle size or the solid-to-liquid ratio to evaluate the impact of surface area.
In the literature, other factors that may affect adsorption include: the size of the solute molecule in relation to adsorbent pore size, the polarity of the contaminant and adsorbent and how likely they are to attract, pH, and solubility (Armenante). Substances that are slightly soluble in water are more easily adsorbed than highly soluble ones because they are less strongly bound with the water (“Solution Chemistry”). MSDS sheets for all three contaminants describe them as soluble in water, but the amount is unspecified. Fertilizer was the least easily adsorbed contaminant, which may be because it is more soluble in water. These factors could be further studied to understand their role in the adsorption process.
One drawback of using a refractometer to measure concentration is its inability to evaluate opaque solutions, as light must pass through to measure IR. Because of this, pesticides, a contaminant of concern in water quality, were not studied in this project.
Conclusions
1. Agricultural wastes are effective to use as adsorbent materials for removing contaminants from water.
2. Chemical contaminants are effectively removed from water by batch adsorption.
3. Contact time is an important factor in batch adsorption, and there is an optimal length of time for mixing that allows the greatest reduction of contaminant concentration.
4. Using the index of refraction (IR) to measure solution concentration is a reliable way to collect and calculate data before and after batch adsorption.
Relevance
Information learned in this project demonstrates the potential for using alternative materials to improve water quality. Results showed that using agricultural waste in a batch adsorption process can remove contaminants from water. With continued research, this process could be used to design a system for small communities or private homes with wells, helping areas without access to public water treatment, including developing countries. More research could be done to increase the effectiveness of agricultural wastes as adsorbents, with the goal of replacing costly activated carbon in public treatment systems.
Local agricultural wastes provide a readily available, low-cost option for adsorbent materials. This benefits the environment by decreasing disposable waste, and may offer a recyclable material for use as compost with slow-release chemical agents. For example, solids recovered from batch adsorption could be applied to fields or landscaping for fertilization, pest or weed control. Because agricultural wastes are organic, they would decompose and provide compost, as well as reuse the contaminant for its stated purpose. As shown in this project, contaminants can be unbound from adsorbents; I hypothesize that is what would happen with water irrigation and time. It would be great to study the effectiveness of the chemicals in this type of recycled product. Other researchers suggest using adsorbents as barriers around crops to stop agricultural runoff from flowing toward surface water and seeping into groundwater (Bakouri et al. 336). These applications have the goal of proactively reducing contamination. Presenting new information and methods to the agricultural community improves awareness about environmental protection and management, and educates about practical solutions.
Bibliography
“About Activated Carbon Filtration.” APS TechNotes. APS Water, 2004. Web. 13 Feb. 2013.
“Adsorption/Active Carbon.” Water Treatment Solutions. Lenntech 2011. Web. 27 March 2012.
“Agricultural Chemicals: Occurrence and Fate of Chemicals Used for Agriculture.” Toxic Substances Hydrology Program. U.S. Geological Survey, 2010. Web. 21 March 2012.
Armenante, Piero. Chemistry 685-Adsorption, New Jersey Institute of Technology, 1999.Web. 16 Sept. 2012.
Artiola, Janick, Kathryn Farrell-Poe, and Jacqueline Moxley. Arizona: Know Your Water. Tucson: University of Arizona, 2004.
Bakouri, Hicham, et al. “Potential Use of Organic Waste Substances as an Ecological Technique to Reduce Pesticide Groundwater Contamination.” Journal of Hydrology 353 (2008): 335-342.
Biello, David. “Fertilizer Runoff Overwhelms Streams and Rivers, Creating Vast Dead Zones.” Scientific American, n.d. Web. 6 March 2011.
Bungay, P. Chemical Engineering—Adsorption. Rensselaer Polytechnic University, 2000. Web. 26 March 2012.
Chang, Ken-Lin, Jun-Hong Lin, and Shui-Tein Chen. “Adsorption Studies on the Removal of Pesticides (Carbofuran) Using Activated Carbon from Rice Straw Agricultural Waste.” World Academy of Science, Engineering and Technology 76 (2011): 348-351.
“Drinking Water Contaminants.” National Primary Drinking Water Regulations. U.S. Environmental Protection Agency, 2012. Web. 30 Oct. 2012.
“Managing Stormwater Runoff to Prevent Contamination of Drinking Water.” Source Water Protection Practices Bulletin. U.S. Environmental Protection Agency, 2010. Web. 18 March 2012.
“Millennium Development Goal: Drinking Water Target Met.” Media Centre. World Health Organization, 2012. Web. 20 March 2012.
“Private Ground Water Wells.” Private Water Systems. Centers for Disease Control and Prevention, 2011.Web. 18 March 2012.
“Refractive Index as a Concentration Measurement Principle, Technical Note 1.00.05.” K-Patents Process Instruments. K-Patents, 2011. Web. 16 Sept. 2012.
“Safe Drinking Water Act.” Safe Drinking Water. United States Environmental Protection Agency, 2012. Web. 18 March 2012.
“Solution Chemistry.” New World Encyclopedia Online. New World Encyclopedia, 2006. Web. 7 Apr. 2013.
“Study Confirms Glyphosate Contaminating Groundwater.” GMWatch, n.p.,2011. Web. 21 March 2012.
Tshabalala, Mandla. “Use of Modified Pine Bark for Removal of Pesticides from Stormwater
Runoff.” The North American Surface Water Quality Conference & Exposition. July 28-31, 2003, San Antonio, Texas. Santa Barbara, CA: Forrester Communications, Inc., 2003.
“Understudied Fungicides Common in U.S. Streams Draining Agricultural Land.” Toxic Substances Hydrology Program. U.S. Geological Survey, 2011. Web. 21 March 2012.
“Water for Life, 2005-2015.” International Decade for Action. United Nations, 2012. Web. 20 Oct. 2012.
“Water Treatment.” Public Water Systems. Centers for Disease Control and Prevention, 2009. Web. 11 March 2012.